Post on 12-Jul-2020
10.1 THE WINTER WEATHER W ARNING DECISION M AKING WO RKSH OPS
Bradford N. Grant*, Edward Mahoney, Andrew Wood, James LaDue, and Xuechao Yu
NOAA/NWS/WDTB
Norman, OK
1. INTRODUCTION
The Warning Decision Training Branch
(WDTB) delivered four training workshops on winter
weather Warning Decision Making (WDM ) during a
two year span from October 2001 to July 2003. The
primary objective for this three and a half day
workshop was to deliver instruction to National
Weather Service (NWS) forecasters on various winter
weather forecasting and warning concepts, techniques
and strategies. The aim was to help forecasters improve
their ability to successfully apply these techniques in the
forecast process and ultimately, provide better service
with more timely and effective winter weather watches,
warnings, and advisories.
This paper provides a look at some of the
unique instructional and operationally oriented
techniques of the winter weather workshop. Being
learner-centered and performance-based was the goal
for this workshop. WDTB instructors strove to establish
this important relationship: each topic or presentation
needed to be verifiable in the laboratory (i.e., concepts
and/or techniques must be applicable in the 3-day
simulation). This relationship is a proven formula for
ensuring success in training (Stolovitch and Keeps,
2003).
The workshop objectives were developed from
the specific job task Knowledge, Skills, and Abilities
(KSAs) that a team of Subject Matter Experts (SMEs)
developed from the Winter Weather Professional
Development Series (PDS)
(http://www.wdtb.noaa.gov/workshop/WinterWx/index
pds.html)
2. WORKSH OP FORMAT
We structured the format of the workshop
presentations to mesh with the primary objectives and
forecast methodology utilized for the 3-day laboratory
Weather Event Simulator (WES) winter storm
simulation (see section 3). This meant that the daily
instructional presentations (which were typically
conducted in the morning) were highly related to the
learning objectives for latter that same day in the
* Corresponding author address: Brad N. Grant,
WDTB, 3200 Marshall Ave., Suite 202, Norman, OK,
73072.
laboratory simulation. Thus, on the first day, the focus
of the presentations was on the watch phase, the second
day on the warning phase, and the third day,
nowcasting.
For the first day, following introductions and
statements of workshop objectives, a representative
from the NW S Office of Climate, Water, and Weather
Services (OCW WS) briefed the class on winter weather
products and services. T his discussion was especially
timely as an updated NWS directive on Weather
Forecast Office (W FO) winter weather products
specifications was just being publicized.
Next, we had presentations aimed at
climatology (forecasting the rarity of storms), ensemble
forecasting (recognizing means and spreads), synoptic
scale forecasting of precipitation type, and numerical
weather prediction in general. Since the models are the
primary tools in preparing the watch product, we
stressed recognizing model signals at this stage. The
winter weather watch products were defined in the
workshop to give a 12 to 48 hour advance no tice of a
hazardous winter weather event which has the potential
to threaten life or property.
The focus of the second day was the warning
phase. Warning thresholds were defined as a situation
when hazardous winter weather is occurring, imminent,
or has a high probability of occurrence during the 12 to
36 hour period. So, we concentrated on how to analyze
model forecast soundings using various techniques such
as the top-down approach (Baumgardt, 1999) for
forecasting precipitation type (12 to 36 hours out) .
Plus, we introduced the use of BUFKIT (Mahoney,
2000) to visualize hourly model sounding parameters
such as dendritic snow growth zones, thermal profiles,
omega, and quantitative precipitation forecasts (QPF).
BUFKIT profiles were generated from running
the Workstation Eta (WS Eta) model. Multiple runs
over each domain used in the simulation were available
for bo th the second and third days of the workshop. W S
Eta model fields were displayable on AWIPS D2D (see
figure 1) or with the BUFKIT sounding analysis
program (see figure 2). BUFK IT was used to display
hourly forecast profiles for multiple sites.
We also had presentations on the second day
of the workshop on diagnosing the effects of
frontogenesis and associated mesoscale banding
phenomena in winter storm events. This was a very
popular presentation as the concepts contained a
technique that would improve operational forecasting
Figure 1. WES display of WS Eta 700 mb omega 9 hr
forecast (contour interval every 2 µbar/sec) and 0.5
deg. radar reflectivity mosaic (shaded) valid 0900
UT C 11 December 2000 .
Figure 2. BUFKIT overview display from the W S Eta
for Valparaiso (VPZ), IN for the simulation event.
Hourly QPF with post-processed precipitation type
shaded in bar graph format along x-axis (darker
shades=mixed precip), negative omega (2µbar/s solid
contoured), dendritic growth zone contoured in thick
lines.
Figure 3. Depiction of collaboration flow of nested
CWAs used in the winter weather simulation. The three-
letter station identifiers are as follows: ILX (Lincoln ,
IL), LOT (Chicago, IL), IND (Indianapolis, IN), ARX
(La Crosse, WI), IWX (North Webster, IN), DMX (Des
Moines, IA), MKX (Milwaukee, WI), LSX (St. Louis,
MO), HPC (Hydrometeorological Prediction Center).
of heavy snow bands.
The third day (nowcast phase) we showed
techniques for how to augment model forecasts with
real-time integrated sensor data.
On Friday (last day of workshop) we wrapped
up the exercise by showing the verification of the event,
plus we discussed some lessons learned.
3. SIMULATION FORMAT
We first had to select a good winter storm case
that would present a forecasting challenge and offer
ample opportunities for training across multiple regions
in the United States. We choose a case that occurred
during 9-11 December 2000 which affected portions of
the Central Plains, Upper Midwest and Great Lakes
regions in the United States. Despite a great case, using
a three-day case posed some problems for the
workshop team.
When designing the case simulation, the key
goal was to relate the simulation into as much as
possible the material presented in the lectures each day.
With an agenda filled with presentations on specific
job-related skills, the three-day simulation concept was
a natural extension of that goal. There were some key
issues that needed to be overcome, however, to make
the simulation work in the proposed manner. The rest
of this section will focus on these issues and the results
of overcoming them.
The idea of using multiple County Warning
Areas (CW As) for the simulation was unprecedented. In
previous workshops, participants “worked” the
simulations as one or at most, two CWAs throughout
the entire case (for severe convective events, the
workshop simulations usually last an hour or two). The
winter weather workshop, on the other hand, utilized a
totally new concept. Our simulation configuration was
three forecasters working as a team at nine different
workstations. The teams issued simulated products for
Figure 4. The WWA simulator tool used in the
workshop to display watches, warnings, and advisories.
Figure 5. WESSL “pop-up” window used in the
simulation.
eight different CWAs and the Hydrometeorological
Prediction Center (HPC) (see figure 3), whose team was
responsible for issuing national guidance products. Each
team worked the same event over a three day period.
The AW IPS D2D software is not necessarily designed
to work in this manner, especially when CW A-specific
data were utilized in the simulation. Thus, the devised
solution led to some creative nesting of the data
directories in D2D.
A second issue that came up conducting
simulation was the local data generated for each CW A.
At the time these workshops were first implemented,
use of on-station models, such as the W S Eta, was still a
new concept. It took some time to test the model
parameterizations to get output that would provide
added value to the operational models and be useful for
applying the forecasting techniques discussed during the
classroom presentations. Once the model
parameterizations were set, then all of the model runs
for each of the eight CWAs were created, which took
longer than anticipated because of file size issues. Once
the data were ready in their final form, the process to
get the data viewable in AWIPS was a big challenge
due to a lack of expertise.
A third significant issue dealt with the
unavailability of a Watch/Warning/Advisory (WW A)
tool for generating and displaying products in the
simulation. Since collaboration among offices was a
key concept of the workshops (and the simulations), a
substitute program was created. T his program, which is
referred to as the W WA Simulator, a llowed participants
to issue products via W arnGen and track the products
on a G raphical User Interface (G UI) to see how their
products matched up with surrounding offices (see
figure 4). This tool was invaluable during the workshop
as a way of visualizing CWA collaboration and
improving temporal and spatial consistency of watches
and warnings.
By overcoming these three issues, the
workshop experience was enhanced. The workshop
participants were able to represent the different CWAs
(and national center) with only a few minor technical
issues. Having multiple CWAs allowed the participants
to deal with a variety of forecast issues that could not
have been represented using only a single CWA. The
participants were able to use the Weather Event
Simulator (WES) technology with AWIPS to utilize
new forecast techniques and data sources. One of the
new features used in WES was the Weather Event
Simulator Scripting Language (WESSL). This tool
provided participants with simulated real-time queries,
spotter reports and messages from emergency managers,
and decision makers (See figure 5).
The daily “blast-up” calls headed by the HPC
team occurred each day during the simulation where the
class could discuss the event and come up with a
common “big picture” of the upcoming event. The
coordination and collaboration process simulated some
aspects of operational HPC winter weather activities
and led to more consistent products being issued by
neighboring CWAs. If there were any product
discrepancies, the participants utilized the WWA
simulator to reso lve product differences.
4. SUMM ARY
The winter weather workshop utilized some
innovative techniques to help students visualize and
practice techniques for winter weather forecasting.
Incorporating techniques such as use of BUFKIT for
precipitation type forecasting into the simulation led to
greater usability and understanding of key concepts
presented at the workshop . The training objectives
were met, which ultimately can lead to more effective
winter weather products and services from the NWS.
5. REFERENCES
Baumgardt, Dan, 1999: Precipitation type forecasting,
A VISITview teletraining presentation
available online at
http://www.cira.colostate.edu/ramm/visit/ptype
.html.
Mahoney, E. A., and T. A. Niziol, 2000: BUFKIT : A
software application toolkit for predicting lake
effect snow. Preprints, 13th Intl. Conf. On
Interactive Info. and Processing Sys. (IIPS) for
Meteorology, Oceanography, and Hydrology,
Amer. Meteor. Soc., Long Beach, CA.
Stolovitch, Harold D, and Erica J. Keeps, 2003: Telling
ain’t training. American Society for Training
Development (ASTD), Alexandria, VA, 193
pp.
The Warning Decision Training Branch (WDTB)
Winter Weather W orkshop , the complete list
of presentations are available online at
http://wdtb.noaa.gov/workshop/W interW xIV/i
ndex.html.